You are using an older browser, so things might be a little slow to load for you and may not look perfect when they do with minor functionality degradation. We suggest you upgrade to a modern browser if you can.
This site uses some features which aren’t supported by your web browser. Some parts of the site will not work correctly. Please consider upgrading to a more modern browser.

About

How We Think About Service Improvement

Our methodology is based on a statistical process control (SPC) rule set. This is the maths behind a number of quality control and service improvement techniques including Lean and Six Sigma methodologies. It describes the variation in a system, with the statistical demonstration of whether this variation is meaningful, or not.

The aim of our approach is not to ‘name and shame’ but to indicate difference. It is then up to the user to ask the right questions: what does this mean? Is this difference a concern? What do I need to do?

By informing the user that the difference exists and is statistically meaningful our system provides vital initial steps in the process that are currently not easily available to most organisations.

There are a number of reasons why we call Stethoscope a service improvement tool rather than a performance management tool. We make available a view of all organisations nationally both in the public view and the more granular subscriber view. This means that for a given indicator or sub-indicator users can see which organisations are high performing for the item in focus, benchmarking against other organisations in England.

Different time periods are sometimes used for different indicators, which initially can be confusing for users, however this allows the most recent data available to be displayed for any given indicator. Some indicators are potentially susceptible to rapid change, such as readmission rates which are shown as single monthly snapshots, whereas others such as mortality rates are shown as ‘year to latest date’.

Finally we use the unadjusted control limit methodology rather than adjusting for ‘over dispersion’. Health care data is commonly quite widely dispersed and for performance management purposes it is useful to adjust for this, we have chosen not to as this is a more sensitive approach and should alert users earlier to areas that need consideration and potentially action.